Minimum regulation of uncoordinated matchings

TL;DR

This paper investigates a game-theoretic approach to forming maximum matchings in a graph by minimally fixing some players' strategies, balancing the costs of intervention against the goal of social optimality.

Contribution

It introduces the problem of minimal strategic fixing to ensure maximum matchings and provides NP-hardness proof along with a constant ratio approximation algorithm.

Findings

The problem of minimal fixing for maximum matchings is NP-hard.

A constant ratio approximation algorithm is proposed for the problem.

Strategic fixing can effectively lead to social optimality in uncoordinated matchings.

Abstract

Due to the lack of coordination, it is unlikely that the selfish players of a strategic game reach a socially good state. A possible way to cope with selfishness is to compute a desired outcome (if it is tractable) and impose it. However this answer is often inappropriate because compelling an agent can be costly, unpopular or just hard to implement. Since both situations (no coordination and full coordination) show opposite advantages and drawbacks, it is natural to study possible tradeoffs. In this paper we study a strategic game where the nodes of a simple graph G are independent agents who try to form pairs: e.g. jobs and applicants, tennis players for a match, etc. In many instances of the game, a Nash equilibrium significantly deviates from a social optimum. We analyze a scenario where we fix the strategy of some players; the other players are free to make their choice. The goal is to compel a minimum number of players and guarantee that any possible equilibrium of the modified game is a social optimum, i.e. created pairs must form a maximum matching of G. We mainly show that this intriguing problem is NP-hard and propose an approximation algorithm with a constant ratio.